[Skip to Navigation]
Sign In

Featured Clinical Reviews

January 26, 2016

Recommendations for Screening for Depression in Adults

Author Affiliations
  • 1Perelman School of Medicine, Department of Psychiatry, University of Pennsylvania, Philadelphia
  • 2Corporal Michael J. Crescenz VAMC, Philadelphia, Pennsylvania
JAMA. 2016;315(4):349-350. doi:10.1001/jama.2015.18406

Although major depressive disorder (MDD) is one of the world’s great public health problems, the morbidity and increased mortality associated with this common illness can be attenuated by the large number of effective treatments that are now widely available. It is therefore important to ensure that efficient methods for population screening are in place and directly linked to health care systems so depressed patients receive appropriate treatment. This is particularly important because effective treatments of depression not only reduce symptoms associated with the disease and reduce the risk of suicide, but also can improve functioning and offset the negative effects that depressive symptoms can have on physical well-being. In the United States, like many other economically developed nations, primary care practices represent the best place for implementation of these methods, because it is the only venue where both screening and, if clinically indicated, treatment can be provided. With this ideal in mind, in 20021 and 20092 the US Preventive Services Task Force (USPSTF) published systematic reviews and recommendations for such screening for adults. In this issue of JAMA, Siu and colleagues3 provide the second update on this important topic.

As the field has evolved, some aspects have changed since publication of the USPSTF’s first report in 2002. Nevertheless, the 2 most essential recommendations are unchanged: (1) sufficiently reliable patient-reported screening scales for depression are readily available and feasible in primary care, and (2) such screening can lead to accurate diagnosis and treatment in primary care. With respect to specific screening tools, the 9-item depression scale of the Patient Health Questionnaire (PHQ-9) has emerged as the first choice for most primary care settings and adult populations; this instrument can accurately identify about 80% of “true” cases, with a positive predictive value of about 50%.3 Clinicians should be aware that such sensitivity, while acceptable for screening, does not mean that a lower score rules out a diagnosis of depression (about 1 in 5 true cases score below the threshold) and that about half the time a “positive” score will result from symptoms that are attributable to other disorders. Thus, as emphasized in the original report,1 screening does not obviate the need for a diagnostic evaluation. The USPSTF concluded that additional work is necessary to confirm the utility of screening in languages other than English and Spanish and notes that another measure, the Edinburgh Postnatal Depression Scale, is better validated for use in pregnant or postpartum populations. Although it is not absolutely necessary to have “1 scale that fits all,” it would not be difficult to ascertain if the PHQ-9 likewise performs well among women who are pregnant or during the immediate postpartum period.

Although the USPSTF report clearly supports routine screening for adults, it acknowledges that the optimal frequency of such screening has not been established. One reason for such uncertainty is that the benefits of screening are directly linked to the probability that a given patient or group of patients will become depressed during a specific time interval. For people with a history of depression, it would make sense to “screen” for illness activity at each visit. For groups at intermediate risk, such as patients receiving regular care for chronic medical conditions such as diabetes or heart disease, it is reasonable to screen at least once each year. For patients in generally good health who only see their primary care physicians sporadically, it may make sense to screen at each visit, although it is likely that a person who rarely sees a physician may not necessarily schedule an appointment to see a primary care physician within weeks or even months of onset of a depressive syndrome. For such individuals, it may more sense to incorporate periodic web-based “health checks.”

One area of clear progress in the field reflected in the current report pertains to the USPSTF’s statement about “…adequate systems… to ensure accurate diagnosis, effective treatment, and appropriate follow-up.”3 When the recommendations were first published in 2002, some of the seminal studies of the ambulatory care models that are now referred to by names such as collaborative care or depression care management were still under way, and it was not yet clear that the outcomes of depression treatment in primary care could be meaningfully enhanced by relatively simple, cost-effective strategies. The current USPSTF report implicitly incorporates the relatively consistent findings from a large number of studies that have been published over the past several decades4 and now asserts that these forms of targeted clinical support have become the standard for ambulatory care. Indeed, the STAR*D study5 showed that patients treated in primary care practices who were provided with a moderate level of clinical support had the same chances for response and remission as patients treated in psychiatric practices.

Although progress has been made, there is still much work to be done, as reflected by the USPSTF’s assessment of a B grade for the overall weight of evidence in support of depression screening in primary care. In part, the B grade reflects the nature of MDD, ie, a clinically heterogeneous group of conditions that respond variably to a diverse group of interventions. The symptom severity spectrum of MDD includes both a milder group of depressions that are quite responsive to many interventions and a more persistent and disabling group of depressions that are more likely to be resistant to standard treatments. Within this context, it is widely known that first-line antidepressant medications have relatively modest specific effects in controlled studies of ambulatory populations (ie, intent-to-treat response rates of about 50%),6 and no valid biomarkers are available to help guide the selection of specific treatments.7 Moreover, in controlled studies, the specific effects of antidepressants are typically smaller than the effects that are attributable to placebo expectancy and other nonspecific factors.6,8,9 Thus, for some depressed patients at the milder end of the symptom severity spectrum, prompt diagnosis and expeditious treatment may not appreciably improve on a favorable natural history (ie, a short, self-limiting episode with a high likelihood of spontaneous remission). For some patients at the more severe end of the severity spectrum, first-line therapies often are not effective. Being better able to match patients to particular treatments could significantly improve the performance of an efficient screening system.

When clinicians consider what to do for a given patient, the results of studies that use a placebo control condition should not be conflated with those that use no treatment or watchful waiting conditions. Patients being cared for by physicians who are prescribing treatments that are expected to work have a better chance of benefitting than individuals who are enrolled in a clinical trial and receive a double-blind placebo, who in turn have at least twice the chance of remitting as a study participant who is allocated to a waiting list.6 Thus, the decision to treat includes both the modest specific benefit of the intervention and the substantial nonspecific benefits associated with receiving care.

It also may be true that patients respond better to treatments that are matched to their preferences and that preference may matter more for patients who have more severe and persistent depressive episodes.10 In this regard, the USPSTF’s updated recommendations, which emphasize both newer-generation antidepressants and empirically supported psychotherapies, either singly or in combination, make good clinical sense and are fully consistent with contemporary North American practice guidelines.11,12 Care systems also need to take into account that some patients will not respond to multiple adequate trials of standard antidepressants or psychotherapies and may require sequential trials of alternate or more complex treatment regimens.13

Until there are better methods to match patients with specific forms of treatment, the best hope to improve on a B grade for patients with depression may be to adapt care systems to respond more flexibly and decisively to key events that are associated with nonadherence or treatment failure. For example, if the clinicians working within a collaborative care model could rapidly incorporate the information that an initial prescription was not filled or was not refilled, it may be possible to diminish the chances that nonadherence will compromise treatment outcome. Likewise, given evidence that nonresponse is predicted by a lack of symptom improvement during the first 14 days of therapy,14 web-based monitoring of symptoms early in the course of therapy may enable physicians and other mental health professionals to intervene more rapidly and reduce the chances of treatment failure. The same approach to ongoing care could be used to facilitate a more timely transition through treatment algorithms13 and more expeditious referral to specialty care.

Editorials represent the opinions of the authors and JAMA and not those of the American Medical Association.
Back to top
Article Information

Corresponding Author: Michael E. Thase, MD, Department of Psychiatry, University of Pennsylvania, 3535 Market St, Philadelphia, PA 19104 (thase@mail.med.upenn.edu).

Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Thase reported having received grants from Eli Lilly, Forest Laboratories, and Otsula and royalties from American Psychiatric Foundation, Guilford Publications, Herald House, W. W. Norton & Company; his spouse being employed by Peloton Advantage; and having received other support from Alkermes, AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Forest Laboratories, Gerson Lehman Group, GlaxoSmithKline, Guidepoint Global, H. Lundbeck A/S, MedAvante, Merck, Neuronetics, Ortho-McNeil Pharmaceuticals, Otsuka, Pfizer, Roche, Shire US, Sunovion Pharmaceuticals, Takeda, Cerecor, Moksha8, Pamlab (Nestle), Allergan, Trius Therapeutical, and Fabre-Kramer Pharmaceuticals.

Pignone  MP, Gaynes  BN, Rushton  JL,  et al.  Screening for depression in adults: a summary of the evidence for the US Preventive Services Task Force .  Ann Intern Med. 2002;136(10):765-776.PubMedGoogle ScholarCrossref
US Preventive Services Task Force.  Screening for depression in adults: US Preventive Services Task Force Recommendation Statement .  Ann Intern Med. 2009;151(11):784-792.PubMedGoogle ScholarCrossref
Siu  AL; US Preventive Services Task Force.  Screening for depression in adults: US Preventive Services Task Force Recommendation Statement .  JAMA. doi:10.1001/jama.2015.18392.Google Scholar
Coventry  PA, Hudson  JL, Kontopantelis  E,  et al.  Characteristics of effective collaborative care for treatment of depression: a systematic review and meta-regression of 74 randomised controlled trials .  PLoS One. 2014;9(9):e108114.PubMedGoogle ScholarCrossref
Gaynes  BN, Rush  AJ, Trivedi  MH,  et al.  Primary versus specialty care outcomes for depressed outpatients managed with measurement-based care: results from STAR*D .  J Gen Intern Med. 2008;23(5):551-560.PubMedGoogle ScholarCrossref
Thase  ME.  The small specific effects of antidepressants in clinical trials: what do they mean to psychiatrists?  Curr Psychiatry Rep. 2011;13(6):476-482.PubMedGoogle ScholarCrossref
Thase  ME.  Using biomarkers to predict treatment response in major depressive disorder: evidence from past and present studies .  Dialogues Clin Neurosci. 2014;16(4):539-544.PubMedGoogle Scholar
Thase  ME, Larsen  KG, Kennedy  SH.  Assessing the “true” effect of active antidepressant therapy v placebo in major depressive disorder: use of a mixture model .  Br J Psychiatry. 2011;199(6):501-507.PubMedGoogle ScholarCrossref
Fournier  JC, DeRubeis  RJ, Hollon  SD,  et al.  Antidepressant drug effects and depression severity: a patient-level meta-analysis .  JAMA. 2010;303(1):47-53.PubMedGoogle ScholarCrossref
Steidtmann  D, Manber  R, Arnow  BA,  et al.  Patient treatment preference as a predictor of response and attrition in treatment for chronic depression .  Depress Anxiety. 2012;29(10):896-905.PubMedGoogle ScholarCrossref
Gelenberg  AJ.  A review of the current guidelines for depression treatment .  J Clin Psychiatry. 2010;71(7):e15.PubMedGoogle ScholarCrossref
Parikh  SV, Segal  ZV, Grigoriadis  S,  et al; Canadian Network for Mood and Anxiety Treatments (CANMAT).  Canadian Network for Mood and Anxiety Treatments (CANMAT) clinical guidelines for the management of major depressive disorder in adults: II, Psychotherapy alone or in combination with antidepressant medication .  J Affect Disord. 2009;117(suppl 1):S15-S25.PubMedGoogle ScholarCrossref
Rush  AJ, Trivedi  MH, Wisniewski  SR,  et al.  Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report .  Am J Psychiatry. 2006;163(11):1905-1917.PubMedGoogle ScholarCrossref
Szegedi  A, Jansen  WT, van Willigenburg  AP, van der Meulen  E, Stassen  HH, Thase  ME.  Early improvement in the first 2 weeks as a predictor of treatment outcome in patients with major depressive disorder: a meta-analysis including 6562 patients .  J Clin Psychiatry. 2009;70(3):344-353.PubMedGoogle ScholarCrossref